There are two files that contain syntax and six data files.

The .do file "Pronoun models.do" codes the dataset "Pronoun main.dta" and runs the primary analyses as well as the robustness checks. It also is used to generate the data used to construct the figures in the main text as well as the tables contained in the appendix.

The R script "Pronoun figures.R" contains the code to construct Figures 1-3 in the main text as well as the figures in the appendix. There are five .csv files used to produce the figures. The .csv files contain data generated from the .do file as well as several columns used to categorize the data. 

The file "main effects.csv" is used to present results from the experiment in Figure 1. In the file, treatment effects in the column titled "b" along with 95% CIs (in the columns "lower" and "upper") for each of the primary variables (in the column "variable") along with the condition (in the column "condition") as well as a column "party" for partisanship. Finally, the final two columns ("item.order" and "party.order") override the alphabetical order for presentational purposes. 

The remaining .csv files have similar columns. These .csv flies are:
 - "effect by party.csv" used to construct Figure 2
 - "sex and party.csv" used to construct Figure 3
 - "effect by ideology.csv" used to construct the figure in the supplementary material

Finally, we have included a "results.log" which contains all of the results generated from "Pronoun models.do".